Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Open Source Model Profile · facebook

convnext-base-384-22k-1k

convnext-base-384-22k-1k is a ConvNeXt image-classification model from facebook. According to the model card, it is pre-trained on ImageNet-22k and fine-tuned on ImageNet-1k at 384x384.

Publisher
facebook
Task
image-classification
Model type
convnext
License
apache-2.0
Library
transformers
Publication status
Accepted · not indexed

Model overview

convnext-base-384-22k-1k is published by facebook as a ConvNeXt image-classification model. The captured configuration identifies ConvNextForImageClassification with model type convnext. According to the model card, it was pre-trained on ImageNet-22k and fine-tuned on ImageNet-1k at 384x384.

Recorded capabilities

ImageNet-22k to ImageNet-1k at 384x384

According to the model card, the base-sized model was pre-trained on ImageNet-22k and fine-tuned on ImageNet-1k at 384x384.

ConvNet modernized from ResNet

According to the model card, ConvNeXT is a pure ConvNet modernized from a ResNet with Swin Transformer inspiration.

ConvNeXt with Transformers support

The captured configuration reports ConvNextForImageClassification with model type convnext and Transformers library support.

Documented classification workflow

According to the model card, the raw model can classify images into one of the 1,000 ImageNet classes with a documented Transformers workflow.

Use cases in the source record

  • Image classification of inputs into one of the 1,000 ImageNet classes using the documented raw-model workflow.
  • Transformers vision experiments with the documented ConvNextImageProcessor and ConvNextForImageClassification usage pattern.

Limitations and unknowns

  • No parameter count was extracted from this record.
  • No evaluation results were extracted from this record.
  • According to the model card disclaimer, the releasing team did not write the card and the Hugging Face team wrote it instead.
  • Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: facebook/convnext-base-384-22k-1k

Captured: Unknown. Processed: 2026-09-07T19:34:44.386862+00:00.

ConvNeXT (base-sized model) ConvNeXT model pre-trained on ImageNet-22k and fine-tuned on ImageNet-1k at resolution 384x384. It was introduced in the paper A ConvNet for the 2020s by Liu et al. and first released in this repository . Disclaimer: The team releasing ConvNeXT did not write a model card for this model so this model card has been written by the Hugging Face team. Model description ConvNeXT is a pure convolutional model (ConvNet), inspired by the design of Vision Transformers, that claims to outperform them. The authors started from a ResNet and "modernized" its design by taking the Swin Transformer as inspiration. Intende…

F001F002F003F004F005F006F007F008F009F010F011F012F013F014